Merge 2eaf303b63 into c699bafc50

Closes #11537

Authored by: bashonly

test/test_traversal.py

@@ -481,7 +481,7 @@ class TestTraversalHelpers:
             'id': 'name',
             'data': 'content',
             'url': 'url',
-        }, all, {subs_list_to_dict}]) == {
+        }, all, {subs_list_to_dict(lang=None)}]) == {
             'de': [{'url': 'https://example.com/subs/de.ass'}],
             'en': [{'data': 'content'}],
         }, 'subs with mandatory items missing should be filtered'
@@ -507,6 +507,54 @@ class TestTraversalHelpers:
             {'url': 'https://example.com/subs/en1', 'ext': 'ext'},
             {'url': 'https://example.com/subs/en2', 'ext': 'ext'},
         ]}, '`quality` key should sort subtitle list accordingly'
+        assert traverse_obj([
+            {'name': 'de', 'url': 'https://example.com/subs/de.ass'},
+            {'name': 'de'},
+            {'name': 'en', 'content': 'content'},
+            {'url': 'https://example.com/subs/en'},
+        ], [..., {
+            'id': 'name',
+            'url': 'url',
+            'data': 'content',
+        }, all, {subs_list_to_dict(lang='en')}]) == {
+            'de': [{'url': 'https://example.com/subs/de.ass'}],
+            'en': [
+                {'data': 'content'},
+                {'url': 'https://example.com/subs/en'},
+            ],
+        }, 'optionally provided lang should be used if no id available'
+        assert traverse_obj([
+            {'name': 1, 'url': 'https://example.com/subs/de1'},
+            {'name': {}, 'url': 'https://example.com/subs/de2'},
+            {'name': 'de', 'ext': 1, 'url': 'https://example.com/subs/de3'},
+            {'name': 'de', 'ext': {}, 'url': 'https://example.com/subs/de4'},
+        ], [..., {
+            'id': 'name',
+            'url': 'url',
+            'ext': 'ext',
+        }, all, {subs_list_to_dict(lang=None)}]) == {
+            'de': [
+                {'url': 'https://example.com/subs/de3'},
+                {'url': 'https://example.com/subs/de4'},
+            ],
+        }, 'non str types should be ignored for id and ext'
+        assert traverse_obj([
+            {'name': 1, 'url': 'https://example.com/subs/de1'},
+            {'name': {}, 'url': 'https://example.com/subs/de2'},
+            {'name': 'de', 'ext': 1, 'url': 'https://example.com/subs/de3'},
+            {'name': 'de', 'ext': {}, 'url': 'https://example.com/subs/de4'},
+        ], [..., {
+            'id': 'name',
+            'url': 'url',
+            'ext': 'ext',
+        }, all, {subs_list_to_dict(lang='de')}]) == {
+            'de': [
+                {'url': 'https://example.com/subs/de1'},
+                {'url': 'https://example.com/subs/de2'},
+                {'url': 'https://example.com/subs/de3'},
+                {'url': 'https://example.com/subs/de4'},
+            ],
+        }, 'non str types should be replaced by default id'
 
     def test_trim_str(self):
         with pytest.raises(TypeError):
@@ -525,7 +573,7 @@ class TestTraversalHelpers:
     def test_unpack(self):
         assert unpack(lambda *x: ''.join(map(str, x)))([1, 2, 3]) == '123'
         assert unpack(join_nonempty)([1, 2, 3]) == '1-2-3'
-        assert unpack(join_nonempty(delim=' '))([1, 2, 3]) == '1 2 3'
+        assert unpack(join_nonempty, delim=' ')([1, 2, 3]) == '1 2 3'
         with pytest.raises(TypeError):
             unpack(join_nonempty)()
         with pytest.raises(TypeError):

test/test_utils.py

@@ -72,7 +72,6 @@ from yt_dlp.utils import (
     intlist_to_bytes,
     iri_to_uri,
     is_html,
-    join_nonempty,
     js_to_json,
     limit_length,
     locked_file,
@@ -2158,10 +2157,6 @@ Line 1
         assert int_or_none(v=10) == 10, 'keyword passed positional should call function'
         assert int_or_none(scale=0.1)(10) == 100, 'call after partial application should call the function'
 
-        assert callable(join_nonempty(delim=', ')), 'varargs positional should apply partially'
-        assert callable(join_nonempty()), 'varargs positional should apply partially'
-        assert join_nonempty(None, delim=', ') == '', 'passed varargs should call the function'
-
 
 if __name__ == '__main__':
     unittest.main()

yt_dlp/YoutubeDL.py

@@ -4381,7 +4381,9 @@ class YoutubeDL:
             return None
 
         for idx, t in list(enumerate(thumbnails))[::-1]:
-            thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
+            thumb_ext = t.get('ext') or determine_ext(t['url'], 'jpg')
+            if multiple:
+                thumb_ext = f'{t["id"]}.{thumb_ext}'
             thumb_display_id = f'{label} thumbnail {t["id"]}'
             thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
             thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))

yt_dlp/extractor/afreecatv.py

@@ -66,6 +66,14 @@ class AfreecaTVBaseIE(InfoExtractor):
             extensions={'legacy_ssl': True}), display_id,
             'Downloading API JSON', 'Unable to download API JSON')
 
+    @staticmethod
+    def _fixup_thumb(thumb_url):
+        if not url_or_none(thumb_url):
+            return None
+        # Core would determine_ext as 'php' from the url, so we need to provide the real ext
+        # See: https://github.com/yt-dlp/yt-dlp/issues/11537
+        return [{'url': thumb_url, 'ext': 'jpg'}]
+
 
 class AfreecaTVIE(AfreecaTVBaseIE):
     IE_NAME = 'soop'
@@ -155,7 +163,7 @@ class AfreecaTVIE(AfreecaTVBaseIE):
             'uploader': ('writer_nick', {str}),
             'uploader_id': ('bj_id', {str}),
             'duration': ('total_file_duration', {int_or_none(scale=1000)}),
-            'thumbnail': ('thumb', {url_or_none}),
+            'thumbnails': ('thumb', {self._fixup_thumb}),
         })
 
         entries = []
@@ -226,8 +234,7 @@ class AfreecaTVCatchStoryIE(AfreecaTVBaseIE):
 
         return self.playlist_result(self._entries(data), video_id)
 
-    @staticmethod
-    def _entries(data):
+    def _entries(self, data):
         # 'files' is always a list with 1 element
         yield from traverse_obj(data, (
             'data', lambda _, v: v['story_type'] == 'catch',
@@ -238,7 +245,7 @@ class AfreecaTVCatchStoryIE(AfreecaTVBaseIE):
                 'title': ('title', {str}),
                 'uploader': ('writer_nick', {str}),
                 'uploader_id': ('writer_id', {str}),
-                'thumbnail': ('thumb', {url_or_none}),
+                'thumbnails': ('thumb', {self._fixup_thumb}),
                 'timestamp': ('write_timestamp', {int_or_none}),
             }))
 

yt_dlp/extractor/common.py

@@ -279,6 +279,7 @@ class InfoExtractor:
     thumbnails:     A list of dictionaries, with the following entries:
                         * "id" (optional, string) - Thumbnail format ID
                         * "url"
+                        * "ext" (optional, string) - actual image extension if not given in URL
                         * "preference" (optional, int) - quality of the image
                         * "width" (optional, int)
                         * "height" (optional, int)

yt_dlp/extractor/zoom.py

@@ -7,6 +7,7 @@ from ..utils import (
     parse_resolution,
     str_or_none,
     traverse_obj,
+    update_url,
     url_basename,
     urlencode_postdata,
     urljoin,
@@ -34,6 +35,7 @@ class ZoomIE(InfoExtractor):
             'ext': 'mp4',
             'title': 'Prépa AF2023 - Séance 5 du 11 avril - R20/VM/GO',
         },
+        'skip': 'This recording has expired',
     }, {
         # share URL
         'url': 'https://us02web.zoom.us/rec/share/hkUk5Zxcga0nkyNGhVCRfzkA2gX_mzgS3LpTxEEWJz9Y_QpIQ4mZFOUx7KZRZDQA.9LGQBdqmDAYgiZ_8',
@@ -61,41 +63,59 @@ class ZoomIE(InfoExtractor):
         return self._search_json(
             r'window\.__data__\s*=', webpage, 'data', video_id, transform_source=js_to_json)
 
-    def _get_real_webpage(self, url, base_url, video_id, url_type):
-        webpage = self._download_webpage(url, video_id, note=f'Downloading {url_type} webpage')
-        try:
-            form = self._form_hidden_inputs('password_form', webpage)
-        except ExtractorError:
-            return webpage
-
+    def _try_login(self, url, base_url, video_id, form):
+        # This will most likely only work for password-protected meetings
         password = self.get_param('videopassword')
         if not password:
             raise ExtractorError(
                 'This video is protected by a passcode, use the --video-password option', expected=True)
+
         is_meeting = form.get('useWhichPasswd') == 'meeting'
         validation = self._download_json(
-            base_url + 'rec/validate%s_passwd' % ('_meet' if is_meeting else ''),
+            base_url + 'nws/recording/1.0/validate%s-passwd' % ('-meeting' if is_meeting else ''),
             video_id, 'Validating passcode', 'Wrong passcode', data=urlencode_postdata({
-                'id': form[('meet' if is_meeting else 'file') + 'Id'],
+                'id': form[('meeting' if is_meeting else 'file') + '_id'],
                 'passwd': password,
                 'action': form.get('action'),
             }))
+
         if not validation.get('status'):
             raise ExtractorError(validation['errorMessage'], expected=True)
-        return self._download_webpage(url, video_id, note=f'Re-downloading {url_type} webpage')
+
+    def _get_real_webpage(self, url, base_url, video_id, url_type):
+        webpage = self._download_webpage(url, video_id, note=f'Downloading {url_type} webpage')
+
+        data = self._get_page_data(webpage, video_id)
+        if data.get('componentName') != 'need-password':  # not password protected
+            return webpage
+
+        # Password-protected:
+        self._try_login(url, base_url, video_id, form=data)
+        # Return the new HTML document
+        new_url = f"{base_url}rec/share/{data['meeting_id']}"
+        return self._download_webpage(new_url, video_id, note=f'Re-downloading {url_type} webpage')
+
+    def _get_share_redirect_url(self, url, base_url, video_id):
+        """Converts a `/rec/share` url to the corresponding `/rec/play` url, performs login if necessary"""
+        webpage = self._get_real_webpage(url, base_url, video_id, 'share')
+        meeting_id = self._get_page_data(webpage, video_id)['meetingId']
+        redirect_dict = self._download_json(
+            f'{base_url}nws/recording/1.0/play/share-info/{meeting_id}',
+            video_id, note='Downloading share info JSON')['result']
+        redirect_path = redirect_dict.pop('redirectUrl')
+        url = update_url(urljoin(base_url, redirect_path), query_update=redirect_dict)
+
+        if redirect_dict.get('componentName') == 'need-password':
+            # First login, then return redirection URL
+            return self._get_share_redirect_url(url, base_url, video_id)
+
+        return url
 
     def _real_extract(self, url):
         base_url, url_type, video_id = self._match_valid_url(url).group('base_url', 'type', 'id')
-        query = {}
 
         if url_type == 'share':
-            webpage = self._get_real_webpage(url, base_url, video_id, 'share')
-            meeting_id = self._get_page_data(webpage, video_id)['meetingId']
-            redirect_path = self._download_json(
-                f'{base_url}nws/recording/1.0/play/share-info/{meeting_id}',
-                video_id, note='Downloading share info JSON')['result']['redirectUrl']
-            url = urljoin(base_url, redirect_path)
-            query['continueMode'] = 'true'
+            url = self._get_share_redirect_url(url, base_url, video_id)
 
         webpage = self._get_real_webpage(url, base_url, video_id, 'play')
         file_id = self._get_page_data(webpage, video_id)['fileId']
@@ -104,10 +124,12 @@ class ZoomIE(InfoExtractor):
             raise ExtractorError('Unable to extract file ID')
 
         data = self._download_json(
-            f'{base_url}nws/recording/1.0/play/info/{file_id}', video_id, query=query,
+            f'{base_url}nws/recording/1.0/play/info/{file_id}', video_id, query={
+                'continueMode': 'true',  # Makes this return value include interpreter audio information
+            },
             note='Downloading play info JSON')['result']
-
         subtitles = {}
+        # XXX: Would be more appropriate to parse chapters separate from subtitles
         for _type in ('transcript', 'cc', 'chapter'):
             if data.get(f'{_type}Url'):
                 subtitles[_type] = [{
@@ -117,6 +139,19 @@ class ZoomIE(InfoExtractor):
 
         formats = []
 
+        if data.get('interpreterAudioList'):
+            for audio in data.get('interpreterAudioList'):
+                formats.append({
+                    'format_note': f'Intepreter: {audio["languageText"]}',
+                    'url': audio['audioUrl'],
+                    'format_id': f'interpreter-{ audio["icon"].lower()}',
+                    'ext': 'm4a',
+                    # There doesn't seem to be an explicit field for a standardized language code,
+                    # sometimes the `language` field may be more accurate than `icon`
+                    'language': audio['icon'].lower(),
+                    'vcodec': 'none',
+                })
+
         if data.get('viewMp4Url'):
             formats.append({
                 'format_note': 'Camera stream',

yt_dlp/utils/_utils.py

@@ -216,7 +216,7 @@ def partial_application(func):
     sig = inspect.signature(func)
     required_args = [
         param.name for param in sig.parameters.values()
-        if param.kind in (inspect.Parameter.POSITIONAL_ONLY, inspect.Parameter.POSITIONAL_OR_KEYWORD, inspect.Parameter.VAR_POSITIONAL)
+        if param.kind in (inspect.Parameter.POSITIONAL_ONLY, inspect.Parameter.POSITIONAL_OR_KEYWORD)
         if param.default is inspect.Parameter.empty
     ]
 
@@ -4837,7 +4837,6 @@ def number_of_digits(number):
     return len('%d' % number)
 
 
-@partial_application
 def join_nonempty(*values, delim='-', from_dict=None):
     if from_dict is not None:
         values = (traversal.traverse_obj(from_dict, variadic(v)) for v in values)

yt_dlp/utils/traversal.py

@@ -332,14 +332,14 @@ class _RequiredError(ExtractorError):
 
 
 @typing.overload
-def subs_list_to_dict(*, ext: str | None = None) -> collections.abc.Callable[[list[dict]], dict[str, list[dict]]]: ...
+def subs_list_to_dict(*, lang: str | None = 'und', ext: str | None = None) -> collections.abc.Callable[[list[dict]], dict[str, list[dict]]]: ...
 
 
 @typing.overload
-def subs_list_to_dict(subs: list[dict] | None, /, *, ext: str | None = None) -> dict[str, list[dict]]: ...
+def subs_list_to_dict(subs: list[dict] | None, /, *, lang: str | None = 'und', ext: str | None = None) -> dict[str, list[dict]]: ...
 
 
-def subs_list_to_dict(subs: list[dict] | None = None, /, *, ext=None):
+def subs_list_to_dict(subs: list[dict] | None = None, /, *, lang='und', ext=None):
     """
     Convert subtitles from a traversal into a subtitle dict.
     The path should have an `all` immediately before this function.
@@ -352,7 +352,7 @@ def subs_list_to_dict(subs: list[dict] | None = None, /, *, ext=None):
     `quality`  The sort order for each subtitle
     """
     if subs is None:
-        return functools.partial(subs_list_to_dict, ext=ext)
+        return functools.partial(subs_list_to_dict, lang=lang, ext=ext)
 
     result = collections.defaultdict(list)
 
@@ -360,9 +360,15 @@ def subs_list_to_dict(subs: list[dict] | None = None, /, *, ext=None):
         if not url_or_none(sub.get('url')) and not sub.get('data'):
             continue
         sub_id = sub.pop('id', None)
-        if sub_id is None:
+        if not isinstance(sub_id, str):
+            if not lang:
                 continue
-        if ext is not None and not sub.get('ext'):
+            sub_id = lang
+        sub_ext = sub.get('ext')
+        if not isinstance(sub_ext, str):
+            if not ext:
+                sub.pop('ext', None)
+            else:
                 sub['ext'] = ext
         result[sub_id].append(sub)
     result = dict(result)
@@ -452,9 +458,9 @@ def trim_str(*, start=None, end=None):
     return trim
 
 
-def unpack(func):
+def unpack(func, **kwargs):
     @functools.wraps(func)
-    def inner(items, **kwargs):
+    def inner(items):
         return func(*items, **kwargs)
 
     return inner